Supporting information

The PspF-binding domain PspA1-144 and the PspA∙F complex - New insights into the coiled-coil dependent regulationof AAA+ proteins

Osadnik et al. (2015)

contains

Supporting Figures S1-S6Supporting tables S1-S2

N

C

C

N

A

C

CC1 CC2

NTR

coiled-coil core

extended hydro-phobic region

B

CC1 CC2N *

Conservation histogram Proteobacterial PspA

100

0

NTR

CC2

NTR

90°

D

E

NTR

Figure S1

Stereoscopic view of PspA M-domain superimposition C

N

C

N

C

N

C

N

C

N

C

N

C C

N N

Figure S2

WT

E23A

D24A

K27A

L28A

R30A

L31A I33

SQ34

AE35

AM

36A

E37A

D38A

Q136A

Q136K

B

PspA1-144

0

5

10

15

20

25

100

psp

A r

epor

ter

activ

ity (n

orm

aliz

ed)

empty

∆pspA reporter + PspA1-144

γ-proteobacterial PspA

100

0

100

All PspA/IM30 proteins

CC1 CC2N

N-terminal region (NTR) tip region NTR stabilization / FBR

Conservation

putative PspF binding region (FBR)

A

0

1

2

3

4

5

WT

K55A

K55E

R59A

R59E

psp

A r

epor

ter

activ

ity (n

orm

aliz

ed)

C

Regions of conservation in PspA1-144

Preliminary screens for conserved PspF-inhibiting residues in PspA1-144

PspA1-144

D24

L28

K27R30

E37

D38

L31Q34

I33M36

K55

R59

Q136

D Orientation of putative PspF-inhibiting residues in PspA1-144

PspA1-144

∆pspA reporter + PspA1-144

Figure S3

FT W E1 E2 E3 E4 E5 E6 E7

full-lengthPspA E37A

PspF

full-lengthWT-PspA

PspF

psp

A r

epor

ter

activ

ity (M

iller

Uni

ts)

1.3

5.3

1.4

2.8

1-22

21-

222

1-22

2

1-22

21-

144

1-14

41-

144

1-14

4

empty

empty

empty

empty

0

50

100

150

200

250

300

∆pspBWT ∆pspC ∆pspG

Full-length PspA induction effect depends on PspBC

WT

E37A

empty

psp

A r

epor

ter

activ

ity (M

iller

Uni

ts)

0

50

100

150

200

250

300

350

400

450

PspA1-222

Effects of PspA1-222

E37A in a WT reporter

A Effects of PspA1-144

E37A in a WT reporter

psp

A r

epor

ter

activ

ity (M

iller

Uni

ts)

0

50

100

150

200

250

300

350

400

450

WT

E37A

empty

PspA1-144

B

DC

E

PspF concentration is comparable in all tested strains

Co-Purification of full-length PspA and its variant with PspF

WT

E37A

R59E

WT

PspA1-144

PspA1-222

Figure S4

15

45

45

15

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

1:10

2:10

8:10

1:1

1:1 + AMPPNP

1:1+ATP

PspF

3:10

4:10

PspF1-265

1.7:1 + ATP

1.7:1 + AMPPNP

1.7:1

0.17:1

0.35:1

0.5:1

0.7:1

0.85:1

1:1

1.2:1

1.35:1

PspA1-144

PspF265only

1.7:1 + ATP

1.7:1 + AMPPNP

1.7:1

0.17:1

0.35:1

0.5:1

0.7:1

0.85:1

1:1

1.2:1

1.35:1

PspA145

only

Incr

easi

ng c

once

ntra

tion

of P

spA

free PspF

PspA:F complex

nucleotides

A B

0 2 4 6 8 100.00

0.05

0.10

0.15

0.20

0.05 mM ADP

0.25 mM ADP

1 µM PspF

Frac

tion

of U

V a

bso

rptio

n (2

30 n

m)

0.5 mM ADP

app)Sedimentation coef�cient (S

C

Dfree PspA

PspA (µM)0.0 0.2 0.4 0.6 0.8 1.0 2.0

frac

tion

of

fast

sed

imen

ting

spec

ies

(%)

0

20

40

60

80

100

1 µM PspF

WT

E37A

1.8

R59E

2 4 6 8 10 0

0.1

0.2

0.3

10 µM

app)

1µM

PspF6

PspF

Sedimentation coef�cient (S

Frac

tion

of U

V a

bso

rptio

n (2

30 n

m)

PspA·F complex formation in SEC concentration dependence of PspF self-hexamerization complex formation at low PspF concentration

effects of nucleotides at low PspF concentration

0

25

50

75

100

0 500 1000 1500 2000 2500

Pho

spha

te [µ

M]

Time after dilution (s)

Koff (app) ≈ 2.69∙10-4 s-1

Koff determination via de-repression of ATPase activity E

Figure S5

E37/Q136

K55/R59

A Atomistic simulation of PspA∙F complexes

FBR facing PspF

W56

sensor II

0 ns 138 ns

stable interaction

no interactionFBR pointing away

B model of the PspA∙F complex based on simulation

90°

Figure S6

Supplementary Figure Legends

Figure S1: The N-terminal region folds back to the PspF-binding domain. Important amino acids

(in a,b,c): conserved hydrophobic amino acids that attach the NTR to CC1; hydrophobic coiled-

coil core forming residues; laterally protruding hydrophobic amino acids from CC1 and one from

CC2 ( ) that link the hydrophobic core of the coiled coil and the hydrophobic region formed by

the back-folding NTR; a residue of CC2 (Lys128) that contacts the NTR; Asn14 and Ala17,

whose backbone is contacted by Asn14. Asn14 is needed for PspC interaction, see discussion. (a)

front and side view of the back-folding N-terminus.

(b) Schematic depiction of the front view in A, showing the large hydrophobic face of CC1 and

the extended hydrophobic region formed by the coiled-coil core and the back-folding NTR ( ).

(c) Conservation histogram of -proteobacterial PspA (overview) and detailed excerpts showing

the high overall degree of conservation of the highlighted amino acids.

(d) The N-terminal region is flexible, but stably attached to CC1 and CC2 during atomistic

simulations. PspA1-144

molecules (n=19) were aligned in their NTR/CC1 region after 18 ns of

simulation (avg.). Even though the rest of the protein as well as the NTR moved considerably and

the NTR showed high flexibility in the region 13-24 ( ), it was still tightly attached to CC1 ( ).

(e) The less well defined electron density of the crystal structure in the N-terminal region indicates

flexibility of this region as well. Cartoon representation of PspA1-144

, colored according to B-factor

of residues in the crystal structure, arrows as in D.

Figure S2: Stereoscopic views of superimposed backbones of ClpB-MD ( ) and PspA ( ) from

the alignment shown in Fig. 3A, as described in the main text.

Figure S3: Preliminary screen to identify putative change of inhibition variants of PspA1-144.

(a) Full overview of regions of PspA1-144

drawn on top of comparative consensus profiles of -

proteobacterial PspA (upper part) and PspA/IM30-family proteins (lower part) indicating regions

of structurally important amino acids (i.e. conserved in both profiles) as well as amino acids

putatively important for proteobacterial PspAs only (i.e., not conserved in all PspA/IM30-family

proteins). Colors for structurally important amino acids in PspA/IM30 as in Fig.2.

(b, c) Preliminary screen of single amino-acid variants of PspA either in the N-terminal (c) or the

C-terminal (c) part of the region conserved in -proteobacteria using a long-induction LacZ-activity

assay. All data normalized to repression level of the wild-type fragment. Several exchanges show

markedly reduced inhibition of PspF compared to the wild-type fragment ( ), or, in the case of the

R59E variant, slightly increased repression. Variants E37A and R59E were used for detailed

studies of PspF inhibition. Even weaker inhibiting variants (up to 25-fold higher psp levels) still

strongly inhibit PspF, as the empty vector control has a 100-fold increased psp level (panel b,

empty)

(d) Orientation of variants on the PspA structure. Variants from the screen in (b) are colored

depending on the strength of their effect ( - surface residues and non-surface residues with strong

loss of inhibition; weak to no effect on inhibition). Loss of inhibition variants cluster on one face

of PspA and point into the same direction as residues R59 and K55, indicating a putative PspF

binding patch, see also Fig.4A.

Figure S4: Effects of PspA variants in PspF regulation in vivo

(a) The PspF level in reporter strains with constitutively produced PspF is indeed comparable and

not influenced by different variants. Western Blot developed against the strep-tag of PspF-strep

after 3h induction (0.1% arabinose), compare to LacZ assays in Fig.4D.

(b) The upregulating effect of PspA1-144-E37A also exists in the full psp wild-type background and

is therefore independent of pspF overproduction. Long-term induction (compare to Fig.4D, right)

leads to strong induction of psp, while the wild-type fragment shows repression of psp as before.

(c) The E37A variant has the same up-regulating effect if introduced into full-length PspA (PspA1-

222), i.e. upregulation of psp. Experimental setup as in panel b. Note that this upregulation is much

stronger than the general increase of psp induction due to PspA overproduction, which we found

to be largely dependent on PspBC (see panel d).

(d) PspBC cause most of the inducing effect of long-term PspA overproduction. Comparison of

changes in pspA reporter levels in the wild-type psp background (WT) or pspB and pspC single

gene deletion mutants caused by production of either full-length PspA (1-222) or the fragment used

in this study (1-144) in comparison to the empty vector control (empty). The 2.8-fold induction in

the wild-type background is reduced to ~1.4-fold reporter levels in both deletion mutants. The only

PspF interacting fragment PspA1-144 inhibits psp in all backgrounds.

(e) Co-elution of strep-tagged PspF with his-tagged full-length PspA-E37A from the reporter strain,

with wild-type PspA shown as comparison (cf. Fig.4E)

Figure S5: PspA1-144

∙ PspF1-265

complex formation.

(a) Size exclusion chromatography with PspF1-265

and PspA1-144

indicates complex formation.

Ratio of PspA relative to PspF is given on the left at the start of each chromatogram. PspA∙F

appears at very low relative concentrations of PspA (0.17:1), while free PspA only appears at nearly

equimolar concentrations. Preincubation with nucleotides (ATP/AMPPNP) and MgCl2 (second and

third chromatogram from the top) does not change complex formation.

(b) Influence of PspF1-265

concentration on self-hexamerization. Distribution of sedimentation

coefficients of 1 or 10 µM PspF and their relative abundance (integrated absorption at 230 nm).

(c) Relative portion of the fast sedimenting species during titration (see Fig.5B) as percentage of

total UV-absorption (230 nm) in the sample cuvette. Addition of 0.08 µM PspA1-144

already leads

to a shift of 46% of PspF1-265

-absorption (the absorption of PspA is negligible at this point).

(d) Stable oligomerization of PspF1-265

in equilibrium is not promoted by ADP. Distribution of

sedimentation coefficients of 1 µM PspF with 0.05 to 0.5 mM ADP ( 5-fold KD) and their relative

abundance (integrated absorption at 230 nm). Larger concentrations of ADP could not be measured

due the high UV-absorption of nucleotides and the low amount of protein. Since PspF1-265

shows

ATPase-activity in vitro at 1 µM, this data does not exclude transient hexamer formation that occurs

spontaneously. Still, stable oligomerization of PspF upon addition, which could be shown in the

case of PspA (cf. Fig.5B), is not observed.

(e) Determination of the apparent dissociation constant (koff) of the PspA∙F complex. Jump dilution

(1:40) of the concentrated PspA1-144

∙ PspF1-265

complex into the ATPase assay. The koff was

determined (Tummino & Copeland, 2008) via curve fitting, with the hydrolysis rate of an

equilibrated sample serving as reference rate.

Figure S6: Atomistic simulations of complex stability of PspA and PspF.

(a) Four independent atomistic simulations of PspA1-144

and PspF1-275

after 138 ns of simulated time

(starting position of PspA was the same for all of them, shown in white cartoon representation,

results of the four simulations in shades of orange superimposed on the right). Stable complex

formation occurred when the FBR was facing PspF (schematically depicted on the left). In contrast,

PspA diffuses away during simulation when PspA is rotated 180° around its longitudinal axis so

that the FBR points away from PspF initially. In the superimposed pictures of simulations on the

right, all PspFs were omitted for clarity but the two from the starting conditions, since PspF did not

unfold or change its structure considerably during simulations.

(b) The modeled PspA1-144

∙ PspF1-265

complex based on the stable binding conformation shown in

panel b.

Supplemental Table S1. Primers used in this study.

cloning of pspA1-144

pspA-F

CCATGGGTATTTTTTCTCGCTTTGC

pspA144-R

CTCGAGCTGATGACGTAACATCAATGC

single amino acid exchanges

pspA-ex-E23A CTGTTAGAGAAAGCGGCTGATCCACAGAAACTGG /

CCAGTTTCTGTGGATCAGCCGCTTTCTCTAACAG

pspA-ex-D24A GAGAAAGCGGAAGCTCCACAGAAACTGG /

CCAGTTTCTGTGGAGCTTCCGCTTTCTC

pspA-ex-L28A GGAAGATCCACAGAAAGCGGTTCGTCTGATGATC /

GATCATCAGACGAACCGCTTTCTGTGGATCTTCC

pspA-ex-L31A CAGAAACTGGTTCGTGCGATGATCCAGGAGATG /

CATCTCCTGGATCATCGCACGAACCAGTTTCTG

pspA-ex-Q34A CTGGTTCGTCTGATGATCGCAGAGATGGAAGATACACTG /

CAGTGTATCTTCCATCTCTGCGATCATCAGACGAACCAG

pspA-ex-E37A GATGATCCAGGAGATGGCAGATACACTGGTTGAAG /

CTTCAACCAGTGTATCTGCCATCTCCTGGATCATC

pspA-ex-K55A CGTTGGCAGAAAAGGCACAGCTGACTCGC /

GCGAGTCAGCTGTGCCTTTTCTGCCAACG

pspA-ex-K55E CGTTGGCAGAAAAGGAACAGCTGACTCG /

CGAGTCAGCTGTTCCTTTTCTGCCAACG

pspA-ex-R59A GAAAAGAAACAGCTGACTGCCCGTATTGAACAAGCGTC /

GACGCTTGTTCAATACGGGCAGTCAGCTGTTTCTTTTC

pspA-ex-R59E GAAAAGAAACAGCTGACTGAACGTATTGAACAAGCGTCGG /

CCGACGCTTGTTCAATACGTTCAGTCAGCTGTTTCTTTTC

pspA-ex-Q136A CACGCGCTCGCGCCCAGGCATTGATGTTACGTCATCAGTGACTCGAGATATA /

TATATCTCGAGTCACTGATGACGTAACATCAATGCCTGGGCGCGAGCGCGTG

pspA-ex-Q136K CACGCGCTCGCAAACAGGCATTGATGTTACGTCATCAGTGACTCGAGATATA /

TATATCTCGAGTCACTGATGACGTAACATCAATGCCTGTTTGCGAGCGCGTG

BspHI-dependent mutagenesis

pspA-BspHI-L31-F1 TCATGATTCAGGAGATGGAAGATACAC

pspA-BspHI-L31-R1 TCATGAGACGAACCAGTTTCTGTG

pspA-BspHI-R30A TCATGAGTGCAACCAGTTTCTGTGGATCTTC

pspA-BspHI-I33S TCATGAGCCAGGAGATGGAAGATACAC

pspA-BspHI-E35A TCATGATTCAGGCAATGGAAGATACACTGGTTGAAG

pspA-BspHI-M36A TCATGATTCAGGAGGCAGAAGATACACTGGTTGAAGTACG

pspA-BspHI-D38A TCATGATTCAGGAGATGGAAGCAACACTGGTTGAAGTACGTTCTAC

cloning of pspF

pspF-F2 CATATGGCAGAATACAAAGATAATTTACTTGGT

pspF265-R2 GGATCCGGCGATAGCGTCTTCAG

pspF-R3 AAGCTTACTCGAGAATCTGGTGCTTTTTCAACAACG

Primer sequences are given in 5'-3' direction. 1Primers used for introduction of a BspHI site at the position that

codes for Leu31 in the amino acid sequence. 2Restriction sites at the 5' end of the primer are omitted for clarity

(NdeI for pspF-F, BamHI for pspF265-R). 3 Reverse primer (5’ HindIII site) for cloning of pspF into pUL-Ptat

contains a XhoI site before the HindIII site for future introduction of tags

Supplemental Table S2. Data collection and refinement statistics.

Dataset

PspA1-144

native

Se-PspA1-144

MAD remote

Se-PspA1-144

MAD peak

Se-PspA1-144

MAD inflection

Radiation source BESSY BL14.1 BESSY BL 14.1 BESSY BL 14.1 BESSY BL14.1

Wavelength (Å)

Energy (kEV)

0.9184

13.5000

0.9759

12.7100

0.9798

12.6540

0.9799

12.6525

Resolution range (Å)

Highest resolution shell (Å)

39.73 - 1.8

1.9 - 1.8

39.80 - 2.01

2.13 - 2.01

39.87 - 2.01

2.13 - 2.01

39.87 - 2.06

2.18 - 2.06

Space group C2 C2 C2 C2

Cell dimensions

a, b, c (Å) 79.65, 30.38, 80.48 79.72, 30.39, 80.70 79.95, 30.47, 80.87 79.94, 30.43, 80.94

() 90, 115.61, 90 90, 115.27, 90 90, 115.29, 90 90, 115.28, 90

Rmerge 3.1 (39.8) 5.2 (40.9) 4.7 (42.7) 4.8 (40.4)

I / I 20.5 (3.5) 11.0 (2.0) 11.7 (1.9) 12.0 (2.0)

Completeness (%) 96.8 (95.6) 96.8 (94.8) 96.7 (95.1) 96.8 (95.8)

Multiplicity 3.4 (3.4) 2.2 (2.2) 2.2 (2.20) 2.2 (2.2)

Wilson B 37.2 36.8 38.7 40.0

Refinement

No. reflections (work/test) 15170 / 798

Rwork / Rfree 0.213 / 0.264

No. atoms

Protein 1127

Buffer components 8

Water 77

B-factors (Å2)

Protein 54.5

Buffer components 71.8

Water 41.7

R.m.s deviations

Bond lengths (Å) 0.018

Bond angles () 1.87

Ramachandran plot (%)

Favoured

Allowed

Outlier

94.9

2.9

2.2

The values in parentheses are given for the highest resolution shell. For the native dataset Friedel pairs were merged, for the MAD

datasets Friedel pairs were treated as separate reflections.